// "collision" = two records have the same has value so need
//   to be stored at the same position in the hash table. If
//   this happens, hash tables will need to do something
//   clever to recover (like linear probing or chaining),
//   otherwise they will be useless.

// If a hash table has size H, what is the probability that
//   you can add N random records without any collisions?

#include <iostream>

void main()
{ int h;
  cout << "Hash table size? ";
  cin >> h;
  double probok = 1.0;
  for (int n=1; n<=h; n+=1)
  { probok = probok * (h-(n-1)) / h;
    cout << "after inserting " << n << " things, " <<
            "probability all OK = " << probok << "\n"; } }

// We see that the results are hopeless.